From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Message-ID: <53A46827.9000400@arm.com> Date: Fri, 20 Jun 2014 17:58:15 +0100 From: Sudeep Holla MIME-Version: 1.0 Subject: Re: [PATCHv7 2/5] mailbox: Introduce framework for mailbox References: <1402592317-7043-1-git-send-email-jaswinder.singh@linaro.org> <1402592479-7244-1-git-send-email-jaswinder.singh@linaro.org> <53A32927.4060004@arm.com> <53A45C54.2010103@arm.com> In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: quoted-printable To: Jassi Brar Cc: Sudeep Holla , "linux-kernel@vger.kernel.org" , "devicetree@vger.kernel.org" , "mporter@linaro.org" , "patches@linaro.org" , "bjorn@kryo.se" , "ashwin.chaugule@linaro.org" , "gregkh@linuxfoundation.org" , "s-anna@ti.com" , "loic.pallardy@st.com" , "lftan.linux@gmail.com" , "slapdau@yahoo.com.au" , "courtney.cavin@sonymobile.com" , Pawel Moll , "robh+dt@kernel.org" , Mark Rutland , "ijc+devicetree@hellion.org.uk" , "arnd@arndb.de" , "joshc@codeaurora.org" , "linus.walleij@linaro.org" , "galak@codeaurora.org" , "ks.giri@samsung.com" List-ID: On 20/06/14 17:30, Jassi Brar wrote: > On 20 June 2014 21:37, Sudeep Holla wrote: >> >> As I mentioned in the other mail, suppose there's one channel and multip= le >> users >> of it. Each have to acquire the channel, and from then any data received >> will >> be sent to the client. But what is the async data from remote is not for= the >> client currently holding the channel. I also gave the example. >> >> If the remote is the process that controls power managements and allows = you >> to >> set the thermal bounds. If those limits are crossed, it sends any interr= upt >> to >> notify that, while the controller can be busy handling some other >> request(say >> DVFS). >> > Multiple clients of a single mailbox channel asks for a single > 'server' type shim client that simply forwards requests to remote on > behalf of clients but filter out async notifications from replies > coming in from remote. In your case, your 'thermal' client driver will > register itself with the shim server for 'thermal' type async > notifications. The shim server holds the mailbox channel for whole of > its lifetime so you never miss the thermal event. > OK thanks for clarifying this. I too had same understanding, just wanted to= =20 cross check and convince myself the need of shim layer. Regards, Sudeep